In order to optimize utilization of relatively fast storage devices having limited capacity, such as in a tiered file storage system, relatively infrequently accessed files are often compressed into blocks of data and archived in relatively slow storage devices of the tiered storage system. Accessing an archived file, such as in response to a read or write request, generally requires decompressing the entire file, as identified based on a file handle or other identifier included in the request, in order to locate a portion of the file identified by a start offset and/or a length included in the request.
Accordingly, although such read and write requests often identify only a portion of the original file, all of the compressed blocks associated with the original file must be retrieved and decompressed in order to locate the requested portion, in order to service a read request, or locate the portion to be written, in order to service a write request. Additionally, in order to service a write request, the entire original file generally must be recompressed, subsequent to the write procedure, even though only a portion of the original file may have been modified. Retrieving and decompressing all of the compressed blocks associated with a file, and in the case of a write request, recompressing an entire file, requires significant time and resources thereby increasing the response time to the requesting client computing device, which is undesirable.